A method and device for power system restoration zoning considering sub-area structure and scale

Abstract

The present invention provides a power system restoration zoning method and device considering sub-area structure and scale. The method comprises the following steps: abstracting the topology of a power grid to be partitioned, constructing a weighted topology map of the power grid and obtaining an adjacency matrix, establishing a power system zoning model based on the principle of a label propagation algorithm, calculating a label propagation probability matrix, initializing a label classification matrix according to the position of a black start power supply node, introducing a sub-area scale matrix and a historical influence factor into the propagation strategy of the label propagation algorithm, propagating node labels according to the improved label propagation strategy to obtain a label influence matrix, updating the label classification matrix, and simultaneously restoring the label information of the black start power supply node. The node label propagation process is repeated until the label classification matrix converges, and the power system zoning result is output. The power system partitions obtained by the present invention have strong community structure characteristics and similar scales, which reduces the waiting time for interconnection of power system partitions and improves the efficiency of parallel restoration of the power system.

Description

Technical Field

[0001] The present invention belongs to the technical field of power system restoration, and in particular relates to a power system restoration zoning method and device taking into account sub-area structure and scale. Background Art

[0002] Power system failures can be restored in both serial and parallel modes. Large-scale power grids have numerous nodes and complex topologies. To ensure power system reliability, multiple black-start power supplies are required. After a power outage, multiple self-starting black-start power supplies provide initial power for parallel zone recovery. After each zone has independently recovered, they can be synchronized and interconnected, improving the speed and efficiency of recovery across the entire complex, large-scale power grid.

[0003] The parallel partition restoration of the power system includes three parts: restoration sub-area division, sub-area self-recovery, and parallel grid connection operation of the partitions. Among them, the division of the restoration sub-area directly affects the restoration effect of the two stages after the parallel partition restoration. Therefore, a reasonable sub-area division method is crucial for the optimization of the black start power supply layout and even the formulation of subsequent power outage restoration strategies. Its basic principles include: (1) the partition has self-recovery capabilities; (2) each partition is centered on the black start power supply as much as possible; (3) the internal network of the partition is tight, and the connection between partitions is sparse; (4) the number of node lines in each partition is as equal as possible.

[0004] In related technologies, because there are relatively few nodes containing label information in the initial power grid, the early propagation process is highly random and unstable, which can easily cause some label information to propagate endlessly along the path, resulting in a certain label information occupying an excessive proportion in the network, thus triggering the emergence of giant communities. Furthermore, if there are smaller communities near a giant community, these small communities are easily swallowed up, causing the giant community to further expand and trigger a vicious cycle. This contradicts the principle of roughly balanced size in power system restoration zoning. Furthermore, traditional propagation strategies can experience label oscillation, making it difficult for the algorithm to converge and affecting the stability of the zoning method. Summary of the Invention

[0005] In view of the shortcomings of the existing technology, the purpose of the present invention is to provide a power system restoration zoning method and device that takes into account the sub-area structure and scale. A power system zoning model is established, and a community scale matrix is introduced into the label propagation algorithm to limit the expansion of the partition scale. A historical impact factor is introduced to prevent label oscillation and promote convergence. Then, a power system zoning scheme is obtained based on the label propagation algorithm. The power system partition community structure characteristics obtained by the present invention are strong and the scale is close, which is conducive to reducing the waiting time for the interconnection of power system partitions and improving the efficiency of parallel restoration of the power system.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0007] A power system restoration zoning method considering sub-zone structure and scale includes the following steps:

[0008] Step 1: Abstract the grid topology to be partitioned, construct a weighted grid topology map based on the branch recovery time and the active power value flowing through, and obtain the adjacency matrix based on the weighted grid topology map;

[0009] Step 2: Establish a power system partition model based on the label propagation strategy, calculate the label propagation probability matrix, and initialize the label classification matrix according to the black start power node location;

[0010] Step 3: Introduce the sub-area scale matrix and historical impact factor into the propagation strategy of the label propagation algorithm to obtain an improved label propagation strategy. Node labels are propagated according to the improved label propagation strategy to obtain the label influence matrix, ensuring that the sub-area structure has strong characteristics and the scale is as balanced as possible.

[0011] Step 4: Update the label classification matrix according to the label influence matrix, and restore the label information of the black start power node. Repeat the node label propagation process until the label classification matrix converges and output the power system partition result.

[0012] Furthermore, in step 1, the generators and loads in the power grid are extracted as undifferentiated nodes, the lines and transformers are regarded as branches, and the power grid is abstracted into a topological graph G = (V, E), where V represents the node set of the topological graph and E represents the branch set of the topological graph;

[0013] The branch weights are set by considering the branch recovery time and the active power flowing through each branch before the power outage, resulting in a weighted topology. The shorter the branch recovery time, the greater the weight, and the more likely the two nodes connected by the branch are to be classified as the same partition. At the same time, the flow data before the power outage is used to reflect the closeness of the connection between each node in the power grid. The greater the active power value transmitted on the branch, the closer the connection between the two nodes connected by the branch. The calculation formula of the branch weight is:

[0014]

[0015] Where w ij represents the branch weight between node i and node j, |P ij | represents the absolute value of the active power flowing through the branch between node i and node j before the power outage, t ij Represents the recovery time of the branch between node i and node j, * represents the normalization of the value. If there is a branch directly connected between node i and node j, according to P ij , t ijThe size of is used to assign a weight to it, otherwise the weight is 0.

[0016] Furthermore, the adjacency matrix A is used to describe the connection relationship of the power grid topology:

[0017]

[0018]

[0019] Furthermore, the step 2 specifically includes:

[0020] Step 2.1: Update labels based on the edge weights and label information of adjacent nodes. The label propagation strategy is as follows:

[0021]

[0022] In the formula, f(i) represents the label information of node i, C represents the type of label information, N(i) represents the set of adjacent nodes of node i, and w i (f j ) indicates that the target node i and all label information are f j The sum of the edge weights of the adjacent nodes;

[0023] Step 2.2, calculate the label propagation probability matrix:

[0024]

[0025] Where, T ij represents the probability that the label information of node j is propagated to node i, w ij represents the branch weight between node i and node j, w kj represents the branch weight between node k and node j;

[0026] Step 2.3: Set the number of partitions to be consistent with the number of black start power supplies. Assume that there are n nodes and k black start power supplies in the power grid. The node where the black start power supply is located is used as the initial labeled node. Its label information remains unchanged during the propagation process. Assign values to the label classification matrix F in the initial stage, specifically:

[0027]

[0028] The label information of other non-black start power sources and load nodes in the power grid needs to be updated and obtained during the propagation process. Therefore, except for the elements involved in the above formula, the remaining elements are set to zero.

[0029] Furthermore, the step 3 specifically includes:

[0030] Step 3.1: Introduce the sub-area size matrix for dividing the power system restoration zones. The calculation formula for the sub-area size factor is:

[0031]

[0032] Where m c represents the sub-area scale factor of label category c, d c represents the sum of the degrees of all nodes with label category c, and l is the total number of branches in the power grid;

[0033] Calculate the sub-area size matrix, the diagonal elements M of the sub-area size matrix M ij is the corresponding sub-area scale factor value, specifically:

[0034]

[0035] In the formula, the sub-area size matrix M represents the size of the existing sub-areas in the system. The larger the sub-area size of the label, the smaller the corresponding sub-area size impact factor.

[0036] Before each round of label propagation, the sub-area size matrix of the current power grid is calculated and taken into consideration when updating node labels;

[0037] Step 3.2: Introduce historical impact factors;

[0038] Since the diagonal elements of the propagation probability matrix are all zero, it means that the node does not consider its own historical label information when updating the label, which will affect the stability of the algorithm and the partitioning effect;

[0039] Calculate the historical impact factor:

[0040]

[0041] in, represents the historical impact factor of the c-type label of node i at time t, The calculation formula is:

[0042]

[0043] The propagation probability matrix considering the historical influence factors of the nodes:

[0044]

[0045] Step 3.3: Node labels are propagated according to the improved label propagation strategy. The specific label propagation strategy is as follows:

[0046]

[0047] In the formula, f(i) represents the label information of node i, C represents the type of label information, N(i) represents the set of adjacent nodes of node i, and w i (f j) indicates that the target node i and all label information are f j The sum of the edge weights of the adjacent nodes, m(f j ) indicates that the label information is f j The community size impact factor;

[0048] Each node propagates according to the label propagation probability matrix and the sub-area size matrix:

[0049] Y t =T t-1 M

[0050] Where Y is the label influence matrix, Y t is the label influence matrix at time t, which represents the impact of the propagation at time t on the label update of each node. t is the label propagation probability matrix at time t.

[0051] Furthermore, in step 4, the label classification matrix F of this round is updated according to the label influence matrix Y. If Y t The value of the element in the cth column of the i-th row is the largest, then the corresponding label classification matrix F i,c The element is set to 1, and the other elements in the i-th row of the label classification matrix F are set to 0, thereby updating the label information of each node.

[0052] Furthermore, in step 4, the label information of the black start power supply node cannot be changed during the propagation process. After the round of propagation is completed, the value of the black start power supply node in the label classification matrix is restored.

[0053] A power system restoration zoning device considering sub-zone structure and scale, comprising:

[0054] The grid weighted topology acquisition module is used to abstract the grid topology to be partitioned, construct a grid weighted topology map based on the branch recovery time and the active power value flowing through it, and obtain the adjacency matrix based on the grid weighted topology map;

[0055] The power system partition model establishment module is used to establish the power system partition model based on the label propagation algorithm principle, calculate the label propagation probability matrix, and initialize the label classification matrix according to the black start power node location;

[0056] The sub-area size matrix and historical impact factor introduction module is used to introduce the sub-area size matrix and historical impact factor into the propagation strategy of the label propagation algorithm. Node labels are propagated according to the propagation strategy to obtain the label influence matrix, ensuring that the sub-area structure characteristics are strong and the scale is as balanced as possible;

[0057] The power system partition result output module is used to update the label classification matrix according to the label influence matrix, restore the label information of the black start power supply node, repeat the node label propagation process until the label classification matrix converges, and output the power system partition result.

[0058] A device for zoning a power system for restoration taking into account sub-area structure and scale includes a processor and a memory for storing a computer program that can be run on the processor. When the processor is used to run the computer program, it executes the steps of any of the above-mentioned methods for zoning a power system for restoration taking into account sub-area structure and scale.

[0059] A computer storage medium having a computer program stored therein, wherein when the computer program is executed by a processor, the computer program implements the steps of any of the above-mentioned methods for power system restoration zoning that considers sub-area structure and scale.

[0060] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0061] This paper proposes a power system restoration zoning method that considers sub-zone structure and size. Based on the fundamental principles of power system zoning, this method establishes a power system zoning model. A community size matrix is introduced into the label propagation algorithm to limit the expansion of sub-zone size. A historical impact factor is introduced to prevent label oscillation and promote convergence. The zoning scheme is then solved based on the label propagation algorithm. The resulting sub-zone community structure is robust and similar in size, which helps reduce waiting time for sub-zone interconnection and improves the efficiency of parallel power system restoration. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] The accompanying drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the accompanying drawings:

[0063] Figure 1 The present invention is a flow chart of a power system restoration zoning method that takes into account the sub-zone structure and scale.

[0064] Figure 2 This is the result diagram of power system restoration partitioning based on label propagation algorithm.

[0065] Figure 3 This is the power system restoration zoning result diagram considering the sub-area structure and scale of the present invention. DETAILED DESCRIPTION

[0066] The technical solutions in the embodiments of the present invention will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present invention.

[0067] The present invention provides a power system restoration zoning method considering sub-zone structure and scale, comprising the following steps:

[0068] Step 1: Abstract the grid topology to be partitioned, construct a weighted grid topology map based on the branch recovery time and the active power value flowing through, and obtain the adjacency matrix based on the weighted grid topology map;

[0069] Step 2: Establish a power system partition model based on the label propagation strategy, calculate the label propagation probability matrix, and initialize the label classification matrix according to the black start power node location;

[0070] Step 3: Introduce the sub-area scale matrix and historical impact factor into the propagation strategy of the label propagation algorithm to obtain an improved label propagation strategy. Node labels are propagated according to the improved label propagation strategy to obtain the label influence matrix, ensuring that the sub-area structure has strong characteristics and the scale is as balanced as possible.

[0071] Step 4: Update the label classification matrix according to the label influence matrix, and restore the label information of the black start power node. Repeat the node label propagation process until the label classification matrix converges and output the power system partition result.

[0072] This paper proposes a power system restoration zoning method that considers sub-zone structure and size. Based on the fundamental principles of power system zoning, this method establishes a power system zoning model. A community size matrix is introduced into the label propagation algorithm to limit the expansion of sub-zone size. A historical impact factor is introduced to prevent label oscillation and promote convergence. The zoning scheme is then solved based on the label propagation algorithm. The resulting sub-zone community structure is robust and similar in size, which helps reduce waiting time for sub-zone interconnection and improves the efficiency of parallel power system restoration.

[0073] This method was simulated and verified using the New England IEEE-118 node standard calculation example. The IEEE-118 node system consists of 54 generator nodes, 64 load nodes, and 186 branches. Black start power supplies were connected to nodes 1, 55, and 107. The power system was divided into three zones based on the nodes where the black start power supplies were located.

[0074] The selection of the recovery path requires, on the one hand, that the sum of the line charging time and transformer operating time passed by the path be as short as possible, so as to achieve the purpose of quickly starting the non-black start unit and accelerate the recovery process of the power grid; on the other hand, it is necessary to ensure that the recovery path formulated has a sufficiently high recovery success rate to smoothly provide starting power for the non-black start unit. The present invention mainly considers the overvoltage problem of line no-load charging, and uses the capacitance value to measure the success rate of branch recovery. The larger the capacitance value, the greater the reactive power generated by the line no-load charging, which easily leads to excess reactive power and causes overvoltage at the end node of the line, resulting in line recovery failure and affecting the recovery of the unit to be started. Therefore, in the present invention, in the step 1, the generators and loads in the power grid are extracted as indifferent nodes, the lines and transformers are regarded as branches, and the power grid is abstracted as a topological graph G = (V, E), where V represents the node set of the topological graph, and E represents the branch set of the topological graph;

[0075] The branch weights are set by considering the branch recovery time and the active power flowing through each branch before the power outage, resulting in a weighted topology. The shorter the branch recovery time, the greater the weight, and the more likely the two nodes connected by the branch are to be classified as the same partition. At the same time, the flow data before the power outage is used to reflect the closeness of the connection between each node in the power grid. The greater the active power value transmitted on the branch, the closer the connection between the two nodes connected by the branch. The calculation formula of the branch weight is:

[0076]

[0077] Where w ij represents the branch weight between node i and node j, |P ij | represents the absolute value of the active power flowing through the branch between node i and node j before the power outage, t ij Represents the recovery time of the branch between node i and node j, * represents the normalization of the value. If there is a branch directly connected between node i and node j, according to P ij , t ij The size of is used to assign a weight to it, otherwise the weight is 0.

[0078] The connection relationship of the power grid topology graph is described by the adjacency matrix A:

[0079]

[0080]

[0081] The division of power system restoration zones is essentially the classification of nodes within the power grid topology. The Label Propagation Algorithm (LPA), as a typical community discovery theory algorithm, has the characteristics of fast speed, high efficiency, no need for any prior parameters and good classification effect compared with other algorithms. Its low complexity makes it better suitable for large and complex networks. Therefore, this paper divides the power system restoration sub-zones based on the principle of the LPA algorithm. Therefore, step 2 establishes a power system partition model based on the principle of the label propagation algorithm. The label propagation probability matrix measures the probability of label information propagation between nodes, and the label classification matrix records the label classification information of all nodes in the network.

[0082] Specifically, in the present invention, step 2 specifically includes:

[0083] Step 2.1: Update labels based on the edge weights and label information of adjacent nodes. The label propagation strategy is as follows:

[0084]

[0085] In the formula, f(i) represents the label information of node i, C represents the type of label information, N(i) represents the set of adjacent nodes of node i, and w i (f j ) indicates that the target node i and all label information are f j The sum of the edge weights of the adjacent nodes;

[0086] Step 2.2, calculate the label propagation probability matrix:

[0087]

[0088] Where, T ij represents the probability that the label information of node j is propagated to node i, w ij represents the branch weight between node i and node j, w kj represents the branch weight between node k and node j;

[0089] Step 2.3: Set the number of partitions to be consistent with the number of black start power supplies. Assume that there are n nodes and k black start power supplies in the power grid. The node where the black start power supply is located is used as the initial labeled node. Its label information remains unchanged during the propagation process. Assign values to the label classification matrix F in the initial stage, specifically:

[0090]

[0091] The label information of other non-black start power sources and load nodes in the power grid needs to be updated and obtained during the propagation process. Therefore, except for the elements involved in the above formula, all other elements are set to zero.

[0092] Since there are fewer nodes containing label information in the initial network, the early propagation process is highly random and unstable, which makes it easy for some label information to be propagated endlessly along the path, resulting in a certain label information accounting for too large a proportion in the network, thereby inducing the emergence of giant communities. In addition, if there are smaller communities near the giant community, the small community can be easily swallowed up, causing the giant community to expand further, triggering a vicious cycle, which is contrary to the principle that the size of each partition in the power system recovery partition should be roughly balanced. At the same time, traditional propagation strategies may experience label oscillation, making it difficult for the algorithm to converge and affecting the stability of the partitioning method. This patent improves the traditional label propagation algorithm to address the above problems, introduces a community size matrix and a historical impact factor into the propagation strategy, and uses it to divide the power system recovery sub-areas.

[0093] In step 3 of the present invention, the sub-area scale matrix and the historical influence factor are introduced into the propagation strategy of the label propagation algorithm to obtain the label influence matrix, which can ensure that the sub-area structure characteristics are strong and the scale is as balanced as possible, thereby effectively increasing the community modularity and reducing the number of contact lines.

[0094] In the present invention, the step 3 specifically includes:

[0095] Step 3.1: Introduce the sub-area size matrix for dividing the power system restoration zones. The calculation formula for the sub-area size factor is:

[0096]

[0097] Where m c represents the sub-area scale factor of label category c, d c represents the sum of the degrees of all nodes with label category c, and l is the total number of branches in the power grid;

[0098] Calculate the sub-area size matrix, the diagonal elements M of the sub-area size matrix M ij is the corresponding sub-area scale factor value, specifically:

[0099]

[0100] In the formula, the sub-area size matrix M represents the size of the existing sub-areas in the system. The larger the sub-area size of the label, the smaller the corresponding sub-area size impact factor.

[0101] Before each round of label propagation, the sub-area size matrix of the current power grid is calculated and taken into consideration when updating node labels;

[0102] Step 3.2: Introduce historical impact factors;

[0103] Since the diagonal elements of the propagation probability matrix are all zero, it means that the node does not consider its own historical label information when updating the label, which will affect the stability of the algorithm and the partitioning effect;

[0104] Calculate the historical impact factor:

[0105]

[0106] in, represents the historical impact factor of the c-type label of node i at time t, The calculation formula is:

[0107]

[0108] The propagation probability matrix considering the historical influence factors of the nodes:

[0109]

[0110] Step 3.3: Node labels are propagated according to the improved label propagation strategy. The specific label propagation strategy is as follows:

[0111]

[0112] In the formula, f(i) represents the label information of node i, C represents the type of label information, N(i) represents the set of adjacent nodes of node i, and w i (f j ) indicates that the target node i and all label information are f j The sum of the edge weights of the adjacent nodes, m(f j ) indicates that the label information is f j The community size impact factor;

[0113] Each node propagates according to the label propagation probability matrix and the sub-area size matrix:

[0114] Y t =T t-1 M

[0115] Where Y is the label influence matrix, Y t is the label influence matrix at time t, which represents the impact of the propagation at time t on the label update of each node. t is the label propagation probability matrix at time t.

[0116] In step 4, the label classification matrix F of this round is updated according to the label influence matrix Y. If Y t The value of the element in the cth column of the i-th row is the largest, then the corresponding label classification matrix F i,c The element is set to 1, and the other elements in the i-th row of the label classification matrix F are set to 0, thereby updating the label information of each node.

[0117] In step 4, the label information of the black start power supply node cannot be changed during the propagation process. After the round of propagation is completed, the value of the black start power supply node in the label classification matrix is restored.

[0118] In summary, the proposed power system restoration zoning method, which considers sub-area structure and scale, effectively increases community modularity and reduces the number of tie lines. This adheres to the principle of densely packed sub-areas and sparsely packed inter-area connections, facilitating parallel operation of each sub-area. By incorporating a community size factor into the label propagation algorithm, the proposed method can identify sub-areas with similar numbers of nodes and lines and similar sizes, reducing the waiting time for synchronous interconnection between sub-areas and improving restoration efficiency.

[0119] In the embodiment of the present invention, in order to verify the effectiveness of the present invention, the power system is partitioned and restored based on the following methods:

[0120] 1) Method 1: Traditional label propagation algorithm;

[0121] 2) Method 2: The power system restoration zoning method of the present invention considers the sub-area structure and scale, and introduces the sub-area scale matrix and historical impact factor into the propagation strategy of the label propagation algorithm;

[0122] The modularity index is used to measure the quality of power grid partitioning. Its value is the proportion of edges located in the same community minus the expected proportion of edges in the community in a random graph. In actual network topology, its value is usually between 0.3 and 0.7. The closer the modularity value is to 1, the stronger the community structure characteristics of the partition.

[0123]

[0124] Where Q represents the modularity of community division; A represents the unweighted adjacency matrix; l represents the total number of branches in the topology; d i d j They each represent the degree of node i and j; δ represents the community membership of nodes i and j. When nodes i and j are in the same partition, δ(i, j) = 1, otherwise its value is zero.

[0125] The partition results obtained after the two methods converge are as follows Figure 2 and Figure 3 shown.

[0126] The algorithm convergence times and the number of partition connection lines of the two different partition recovery methods are counted, and the corresponding modularity is calculated. The results are shown in the table.

[0127] Table 1 Comparison of the effects of different partition recovery methods

[0128]

[0129] From Table 2 we can see that:

[0130] 1) Compared with method 1, the label propagation algorithm proposed in method 2 of the present invention requires fewer iterations to converge, which shows that introducing the label history influence factor into the algorithm is effective in promoting algorithm convergence and can theoretically better prevent the occurrence of label oscillation.

[0131] 2) The number of partition tie lines in the second method of the present invention is relatively small, which complies with the principle of sparse connections between partitions in power system restoration and is conducive to parallel operation of each partition.

[0132] 3) The modularity of the community partitioning of the present invention is increased by 43.57% compared with the first method, indicating that the second method can greatly improve the strength of the community structure and enhance the closeness of the network connection within the partition.

[0133] Furthermore, the community size factor values for the three partitions obtained by the two methods were compared, and the variance was used to characterize the size differences between the partitions. As shown in the table, the variance of the community size factors obtained by the present invention, due to the introduction of the community size matrix during the label propagation process, was reduced by 61.33%. This indicates that the number of nodes and lines between the partitions is more equal, and the partition sizes are closer. This helps to reduce the synchronization interconnection waiting time after the restoration of each partition and improve the efficiency of parallel restoration of the power system.

[0134] Table 2 Comparison of community size factors

[0135]

[0136] In summary, the power system restoration zoning method proposed in the present invention, which considers the sub-area structure and scale, can obtain restoration sub-areas with strong community structure characteristics and comparable scale, and can be further applied to the subsequent multi-black start power supply layout optimization model.

[0137] A power system restoration zoning device considering sub-zone structure and scale, comprising:

[0138] The grid weighted topology acquisition module is used to abstract the grid topology to be partitioned, construct a grid weighted topology map based on the branch recovery time and the active power value flowing through it, and obtain the adjacency matrix based on the grid weighted topology map;

[0139] The power system partition model establishment module is used to establish the power system partition model based on the label propagation algorithm principle, calculate the label propagation probability matrix, and initialize the label classification matrix according to the black start power node location;

[0140] The sub-area size matrix and historical impact factor introduction module is used to introduce the sub-area size matrix and historical impact factor into the propagation strategy of the label propagation algorithm. Node labels are propagated according to the propagation strategy to obtain the label influence matrix, ensuring that the sub-area structure characteristics are strong and the scale is as balanced as possible;

[0141] The power system partition result output module is used to update the label classification matrix according to the label influence matrix, restore the label information of the black start power supply node, repeat the node label propagation process until the label classification matrix converges, and output the power system partition result.

[0142] A power system restoration zoning device that takes into account sub-area structure and scale, characterized in that it includes a processor and a memory for storing a computer program that can be run on the processor, and when the processor is used to run the computer program, it executes the steps of the power system restoration zoning method that takes into account sub-area structure and scale as described above.

[0143] A computer storage medium having a computer program stored therein, wherein when the computer program is executed by a processor, the computer program implements the steps of any of the above-mentioned methods for power system restoration zoning that considers sub-area structure and scale.

[0144] A device for zoning a power system for restoration taking into account sub-area structure and scale includes a processor and a memory for storing a computer program that can be run on the processor. When the processor is used to run the computer program, it executes the steps of the method for zoning a power system for restoration taking into account sub-area structure and scale as described above.

[0145] The memory in the embodiment of the present invention is used to store various types of data to support the operation of the power system restoration partition device considering the sub-area structure and scale. Examples of such data include: any computer program for operating on the power system restoration partition device considering the sub-area structure and scale.

[0146] The method for zoning a power system for restoration that considers sub-area structure and scale, disclosed in embodiments of the present invention, can be applied to or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the method for zoning a power system for restoration that considers sub-area structure and scale can be performed by hardware integrated logic circuits or software instructions within the processor. The processor may be a general-purpose processor, a digital signal processor (DSP), or other programmable logic device, discrete gate or transistor logic device, discrete hardware component, etc. The processor can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the embodiments of the present invention can be directly implemented and executed by a hardware decoding processor, or by a combination of hardware and software modules within the decoding processor. The software module may be located in a storage medium located in a memory. The processor reads information from the memory and, in conjunction with its hardware, completes the steps of the method for zoning a power system for restoration that considers sub-area structure and scale, provided in the embodiments of the present invention.

[0147] In an exemplary embodiment, the power system restoration zoning device taking into account the sub-zone structure and scale can be implemented by one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), FPGAs, general-purpose processors, controllers, microcontrollers (MCUs), microprocessors, or other electronic components to execute the aforementioned method.

[0148] It is understood that the memory can be volatile memory or non-volatile memory, or can include both volatile and non-volatile memory. Among them, the non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), magnetic random access memory (FRAM), flash memory, magnetic surface memory, optical disk, or compact disc read-only memory (CD-ROM); magnetic surface memory can be magnetic disk memory or tape memory. Volatile memory can be random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static random access memory (SRAM), synchronous static random access memory (SSRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct rambus random access memory (DRRAM). The memory described in the embodiments of the present invention is intended to include, but is not limited to, these and any other suitable types of memory.

[0149] A computer storage medium having a computer program stored therein, wherein when the computer program is executed by a processor, the computer program implements the steps of any of the above-mentioned methods for power system restoration zoning that considers sub-area structure and scale.

[0150] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A power system restoration zoning method considering sub-zone structure and scale, characterized in that: The following steps are involved: Step 1: Abstract the grid topology to be partitioned, construct a weighted grid topology map based on the branch recovery time and the active power value flowing through, and obtain the adjacency matrix based on the weighted grid topology map; Step 2: Establish a power system partition model based on the label propagation strategy, calculate the label propagation probability matrix, and initialize the label classification matrix according to the black start power node location; Step 3: Introduce the sub-area scale matrix and historical impact factor into the propagation strategy of the label propagation algorithm to obtain an improved label propagation strategy. Node labels are propagated according to the improved label propagation strategy to obtain the label influence matrix, ensuring that the sub-area structure has strong characteristics and the scale is as balanced as possible. Step 4: Update the label classification matrix based on the label influence matrix, and restore the label information of the black start power node. Repeat the node label propagation process until the label classification matrix converges and output the power system partition result. The step 2 specifically includes: Step 2.1: Update labels based on the edge weights and label information of adjacent nodes. The label propagation strategy is as follows: Where, Representation node Label information, C represents the type of label information, Representative Node The set of adjacent nodes of Indicates the target node With all label information The sum of the edge weights of the adjacent nodes; Step 2.2, calculate the label propagation probability matrix: Where, Representation node The label information is propagated to the node The probability of Representation node With node The branch weight between Representation node With node The branch weight between Step 2.3, set the number of partitions to be consistent with the number of black start power supplies, assuming that there are nodes, The black start power supply is used as the initial label node. Its label information remains unchanged during the propagation process. The label classification matrix F in the initial stage is assigned a value, which is specifically: The label information of other non-black start power sources and load nodes in the power grid needs to be updated and obtained during the propagation process. Therefore, except for the elements involved in the above formula, the remaining elements are set to zero.

2. The power system restoration zoning method considering sub-zone structure and scale according to claim 1 is characterized by: In step 1, the generators and loads in the power grid are extracted as undifferentiated nodes, the lines and transformers are regarded as branches, and the power grid is abstracted into a topological diagram. , where V represents the node set of the topological graph, and E represents the branch set of the topological graph; The branch weights are set by considering the branch recovery time and the active power flowing through each branch before the power outage, resulting in a weighted topology. The shorter the branch recovery time, the greater the weight, and the more likely the two nodes connected by the branch are to be classified as the same partition. At the same time, the flow data before the power outage is used to reflect the closeness of the connection between each node in the power grid. The greater the active power value transmitted on the branch, the closer the connection between the two nodes connected by the branch. The calculation formula of the branch weight is: Where, Representation node With node The branch weight between Represents the node before the power outage With node The absolute value of the active power flowing through the intermediate branch, Representative Node With node The recovery time of the intermediate branch, * represents the normalization of the value. If the node With node There are directly connected branches, according to 、 The size of is used to assign a weight to it, otherwise the weight is 0.

3. The power system restoration zoning method considering sub-zone structure and scale according to claim 2 is characterized by: Adjacency matrix Describe the connection relationship of the power grid topology: 。 4. The power system restoration zoning method considering sub-zone structure and scale according to claim 1 is characterized by: The step 3 specifically includes: Step 3.1: Introduce the sub-area size matrix for dividing the power system restoration zones. The calculation formula for the sub-area size factor is: Where, Indicates the label category The sub-area size factor, Indicates that the label category is The sum of all node degrees of , is the total number of branches in the power grid; Calculate the sub-area size matrix, the diagonal elements of the sub-area size matrix M is the corresponding sub-area scale factor value, specifically: In the formula, the sub-area size matrix M represents the size of the existing sub-areas in the system. The larger the sub-area size of the label, the smaller the corresponding sub-area size impact factor. Before each round of label propagation, the sub-area size matrix of the current power grid is calculated and taken into consideration when updating node labels; Step 3.2: Introduce historical impact factors; Since the diagonal elements of the propagation probability matrix are all zero, it means that the node does not consider its own historical label information when updating the label, which will affect the stability of the algorithm and the partitioning effect; Calculate the historical impact factor: in, represent Time Node of Class label historical impact factor, The calculation formula is: The propagation probability matrix considering the historical influence factors of the nodes: Step 3.3: Node labels are propagated according to the improved label propagation strategy. The specific label propagation strategy is as follows: Where, Representation node Tag information, Represents the type of label information, Representative Node The set of adjacent nodes of Indicates the target node With all label information The sum of the edge weights of the adjacent nodes, Indicates that the tag information is The community size impact factor; Each node propagates according to the label propagation probability matrix and the sub-area size matrix: Where Y is the label influence matrix, for The label influence matrix at the moment is expressed as The impact of time propagation on the label update of each node, for The label propagation probability matrix at time t.

5. The power system restoration zoning method considering sub-zone structure and scale according to claim 1 is characterized by: In step 4, the label classification matrix F of this round is updated according to the label influence matrix Y. No. No. The column element value is the largest, then the corresponding label classification matrix The element is set to 1, and the label classification matrix F The other elements of the row are set to 0, thereby updating the label information of each node.

6. The power system restoration zoning method considering sub-zone structure and scale according to claim 1 is characterized by: In step 4, the label information of the black start power supply node cannot be changed during the propagation process. After the round of propagation is completed, the value of the black start power supply node in the label classification matrix is restored.

7. A power system restoration zoning device considering sub-zone structure and scale, characterized in that: include: The grid weighted topology acquisition module is used to abstract the grid topology to be partitioned, construct a grid weighted topology map based on the branch recovery time and the active power value flowing through it, and obtain the adjacency matrix based on the grid weighted topology map; The power system partition model establishment module is used to establish the power system partition model based on the label propagation algorithm principle, calculate the label propagation probability matrix, and initialize the label classification matrix according to the black start power node location. Specifically, it includes: Labels are updated based on the edge weights and label information of adjacent nodes. The label propagation strategy is as follows: Where, Representation node Label information, C represents the type of label information, Representative Node The set of adjacent nodes of Indicates the target node With all label information The sum of the edge weights of the adjacent nodes; Calculate the label propagation probability matrix: Where, Representation node The label information is propagated to the node The probability of Representation node With node The branch weight between Representation node With node The branch weight between Set the number of partitions to be consistent with the number of black start power supplies. Assuming that there are nodes, The black start power supply is used as the initial label node. Its label information remains unchanged during the propagation process. The label classification matrix F in the initial stage is assigned a value, which is specifically: The label information of other non-black start power sources and load nodes in the power grid needs to be updated and obtained during the propagation process. Therefore, except for the elements involved in the above formula, all other elements are set to zero. The sub-area size matrix and historical impact factor introduction module is used to introduce the sub-area size matrix and historical impact factor into the propagation strategy of the label propagation algorithm. Node labels are propagated according to the propagation strategy to obtain the label influence matrix, ensuring that the sub-area structure characteristics are strong and the scale is as balanced as possible; The power system partition result output module is used to update the label classification matrix according to the label influence matrix, restore the label information of the black start power supply node, repeat the node label propagation process until the label classification matrix converges, and output the power system partition result.

8. A power system restoration zoning device that takes into account the sub-zone structure and scale, characterized by: The invention comprises a processor and a memory for storing a computer program that can be run on the processor. When the processor is used to run the computer program, the method for restoring and zoning a power system considering the sub-area structure and scale as described in any one of claims 1 to 6 is executed.

9. A computer storage medium, characterized in that: The computer storage medium stores a computer program, characterized in that when the computer program is executed by a processor, the steps of the power system restoration zoning method considering sub-area structure and scale as described in any one of claims 1 to 6 are implemented.

Images